NO313331B1 - O6-alkyl guanine derivatives, process for their preparation and use, and pharmaceutical preparations - Google Patents

Abstract

O &cir& _<6>-hetarylalkyl- or naphthylalkylguanine derivatives of formula (I) wherein Y is H, ribosyl, deoxyribosyl, or R"XCHR"', wherein X is O or S, R'' and R''' are alkyl, or substituted derivatives thereof; R' is H, or alkyl or hydroxyalkyl; R is (i) a cyclic group having at least one 5- or 6-membered heterocyclic ring, optionally with a carbocyclic or heterocyclic ring fused thereto, the or each heterocyclic ring having at least one hereto atom chosen from O, N or S, or a substituted derivative thereof; or (ii) naphthyl or a substituted derivative thereof; and pharmaceutically acceptable salts thereof, exhibit the ability to deplete O &cir& _<6>-alkylguanine-DNA alkyltransferase (ATase) activity. A process for preparation of the compounds is described. The compounds have utility in combination with alkylating agents in the chemotherapeutic treatment of tumour cells.

Description

Technical area

The present invention relates to 0<6->

alkylguanine derivatives, methods for their production and use, as well as pharmaceutical preparations. In particular, it applies to guanine derivatives which have heteroarylalkyl or naphthylalkyl substituents in the 0 -position, as these compounds have the ability to strongly reduce the activity of 0<6 >-alkylguanine-DNA alkyltransferase (ATase) in tumor cells.

Background technology

It has been proposed to use 0<6->alkylguanine derivatives

which has O^-alkylguanine-DNA alkyltransferase-depleting activity

tet to increase the effectiveness of chemotherapeutic alkylating agents used to kill tumor cells. There is increasing evidence that in mammalian cells the toxic and mutagenic effects

ter of alkylating agents to a large extent is a consequence of alkylation in the 0<6-> position of guanine in DNA. The repair of 0<6->alkylguanine is mediated by ATase, a repair protein that acts on the 0<6->alkylated guanine residues by stoichiometric transfer of the alkyl group to a cysteine residue in the active site of the repair protein in an autoinactivation process. The importance of ATase for the protection of cells against the biological

gic effects of alkylating agents have been clearly demonstrated by the transfer and expression of cloned ATase genes or cDNA into ATase-deficient cells: this confers resistance to several different agents, mainly those that methylate or chloroethylate DNA. Although mechanical

but for cell killing with 0<6->methylguanine in ATase-deficient cells is not yet clear, the killing by 0 6-chloroethyl-guanine takes place through DNA cross-link formation between the strands to a cytosine residue on the opposite strand via a cyclic ethane-guanine -intermediate, a process that is prevented by ATase-mediated chloroethyl group removal or complexation.

The use of 0 -methylguanine and 0 -n-butylguanine for

strongly reducing ATase activity has been investigated (Dolan et al., Cancer Res., (1986) 46, p. 4500; Dolan et al., Cancer Chemother. Pharmacol., (1989) 25, p. 103). O<6->benzylguanine-

derivatives have been suggested to strongly reduce ATase activity to render ATase-expressing cells more sensitive to the cytotoxic effects of chloroethylating agents (Mochel et al., . t. M<p>H. r. hpm., 1992, 35, 4486). US Patent 5,091,430 and International Patent Application No. WO 91/13898 Mochel et. al., describes a method for depleting the level of Q-alkylguanine-DNA alkyltransferase in tumor cells in a host which comprises administering to the host an effective amount of a preparation containing Q-benzylated guanine derivatives of the following formula: where Z is hydrogen, or

and Ra is a benzyl group or a substituted benzyl group.

A benzyl group can be substituted in the nrtn, m<p>fa or paxa position with a substituent group such as halogen, nitro, aryl such as phenyl or substituted phenyl, alkyl with 1-4 carbon atoms, alkoxy with 1-4 carbon atoms, alkenyl with up to 4 carbon atoms, alkynyl with up to 4 carbon atoms, amino, mono-alkylamino, dialkylamino, trifluoromethyl, hydroxy, hydroxy-methyl, and SOnR where n is 0, 1, 2 or 3 and R is hydrogen,

alkyl with 1-4 carbon atoms or aryl. Mi-Young Chae et al., J. Med. rthPTTi. r 1994, 37, 342-447 - published after the priority date of the present application - describes tests on cft-

benzylguanine analogues which have increasingly space-demanding substituent groups on the benzene ring or in position 9. Compound No. 6 described therein is Q 2 -(2-pyridylmethyl)guanine, which in this application is called Q -(2-picolyl)guanine. However, in the results and discussion on pages 342-343 of the publication of Chae et al., compound #6 is not highlighted as a compound of interest, but is grouped among "remaining compounds" that "had moderate activity" (page 343 , lines 12 - 15 i

the text). The authors confirm their previous observations (J_ Med. Chem., 1992, 35 4486) that only allyl or benzyl substituents in the 6-position of guanine would effectively inactivate ATase (p. 343, 1. 21 - 23 in the text).

Q CL-benzylguanine has limitations in its use as an ATase inactivator. It is more stable than desirable, resulting in a long survival time in an animal to which it is administered. It has a level of potential toxicity both alone and in combination with chlorethylene agents which is also undesirable and which may relate to survival time.

The compounds according to the present invention have ATase inactivation properties which are different from Q -benzylguanine, and in some cases the activity is up to 8 times higher than the activity of Q -benzylguanine. Different toxicity characteristics and half-lives have also been observed. It is therefore an object of the present invention to provide new compounds which are useful for depleting the ATase activity in order to increase the effects of chemotherapeutic agents such as anti-tumor agents for chloroethylation or methylation.

The present invention therefore relates to cfc-alkylguanine derivatives, characterized by the formula I:

where

Y is H, ribosyl,

R' is H or q-C 6 alkyl,

R is

(i) a cyclic group with one 5- or 6-membered heterocyclic ring, optionally with a 6-membered carbocyclic or 5- or 6-membered heterocyclic ring fused thereto, or wherein the one or each heterocyclic ring has at least one heteroatom selected from 0, N or S, or a substituted derivative thereof with substitution in the heterocyclic ring(s) and/or carbocyclic ring(s) with one or more of halogen, cyano or S0nR' " where R"" is C^-C^ -alkyl and

n = 0, 1 or 2; or

(ii) naphthyl

and pharmaceutically acceptable salts thereof.

Another object of this invention is to provide pharmaceutical preparations characterized in that they comprise a compound according to any of the preceding claims, and a pharmaceutically acceptable excipient containing compounds which are useful for depleting the ATase activity. The present invention further relates to a method for producing a compound of formula I according to claim 1, characterized in that it comprises the reaction of sodium hydride with a solution of RR'CH0H (where R and R<1> are as defined in claim 1) in an organic solvent, addition of 2-amino-N,N,N-trimethyl-1H-purine-6-aminium chloride or 2-amino-6-chloropurine riboside, treatment with a weak acid and ether, and extraction of the desired product. A further object according to the present invention is to provide a compound for depleting the ATase activities in tumor cells.

R may suitably be a 5- or 6-membered heterocyclic ring or a benzo derivative thereof, in which latter case the O-alkylguanine group may be attached to R in either the heterocyclic or benzene ring.

In preferred embodiments, R shall be a 5-membered heterocyclic ring having at least one S atom therein.

Preferably, R is selected from a thiophene ring, a furan ring or a substituted derivative thereof as defined in claim 1.

Alternatively, R must be selected from a thiophene ring, a furan ring and substituted derivatives thereof selected from bromo- and cyano-substituted derivatives thereof.

Examples of compounds according to this invention (together with compounds B.4214 and B.4218 not covered by the present application) are shown in table 1.

Among the compounds in Table 1, compounds B.4214 and B.4218 are not compounds according to this invention. Compound B.4210 (i.e. the compound of formula I where R is 2-pyrid-yl, R<1> is H and Y is H) is not a preferred compound

according to this invention.

Particularly preferred compounds according to the invention include: Q 6 -thenylguanine, Q 6 -(3-thienylmethyl)guanine, Q 6 -piperonylguanine,

Q 6 -furfuryl-guanine, Q 6 -(2-benzo[b]thienylmethyl)guanine, Q 6-(2-benzofuranylmethyl)guanine, Q 6 -(5-thiazolylmethyl)guanine, Q. 6-(bromothenyl)guanine, Q 6 -(5-cyanofurfuryl)guanine and Q 6-(2-naphthylmethyl)guanine.

The most preferred compounds according to the invention are those which inactivate ATase in we believe and/or in mammalian cells and/or tumor xenografts more effectively than Q f.-benzylguanine (BeG) and which sensitize mammalian cells and/or tumor xenografts to the killing or growth-inhibitory effects of nitroso-urea compounds and/or methylating agents more effectively than BeG. Preferred compounds should also, compared to BeG, have reduced toxicity to normal tissues and/or to the whole organism when used in combination with such agents. Preferred compounds should not themselves be toxic or have more than minimal toxicity at the doses required to inactivate ATase nor should any hydrolysis products of a preferred compound that were chemically unstable be toxic. Although the invention is not limited by theory, it may be necessary for preferred compounds to be less stable than BeG so that they undergo spontaneous chemical degradation immediately after achieving maximal inactivation of ATase: in this way one would minimize any effect of metabolic processes that could act on the agent to form toxic compounds. Preferred compounds should be less able to sensitize human bone marrow or other normal cell types to the toxic effects of alkylating agents, so that they would not exacerbate the known toxicity or create new toxicities of these agents

in normal human tissues.

Preferred compounds according to this invention include those which have a relatively low I5Q value in Table 4 herein (e.g.

below 1.0 uM, more specifically below 0.04 uM) and/or having a relatively short half-life in buffer I (representing conditions for an in vitro assay) and/or phosphate buffered saline (PBS) (representing conditions in physiological medium) in Table 4 here (e.g. under 20 hours in buffer I or under 16 hours in PBS).

A relatively short half-life can be regarded as an indicator that a compound according to the invention would be less stable than O -benzylguanine due to the reactivity of RR'CH- and would tend to break down by hydrolysis in physiological medium.

The influence of the group RR'CH- in the compounds of formula I which enables them to act as ATase inhibitors is determined by electronic, steric and physico-chemical factors. Steric factors can be related to the nature of the surroundings of the cysteine receptor site in ATase. Preferably, R' will be H. A secondary carbonate if bonded to O<6> (as in the DL compounds B.4214 or B.4217) has been found to significantly reduce the inactivation activity, probably due to the bulk of the substituent.

Preferably, the cyclic group in R will not have a methyl group in a neighboring position (as in B.4222), although the influence of neighboring substitution is obviously much smaller when R is heterocyclic than in the naphthyl isomers B.4213 and B.4265.

Physico-chemical factors such as stability, solubility and water/lipid distribution are relevant to the selection of compounds for use in vivo, and affect e.g. formulation, absorption and transport. Selection of compounds can also be influenced by their different distribution in different tissues.

One embodiment of the invention provides for the use of a compound according to any one of claims 1-10,

for the preparation of a drug for the depletion of O -

alkylguanine-DNA alkyl-transferase activity in tumor cells.

In a further embodiment, the use of a compound according to any one of claims 1-10 and an alkylating agent provides for the preparation of a one- or two-part pharmaceutical preparation for the treatment of tumor cells.

Accordingly, it is possible to treat tumor cells in a host. It is also possible to treat neoplasms including those known to be sensitive to the action of alkylating agents, e.g. melanomas and gliomas and others whose resistance to treatment with alkylating agents alone can be overcome by the use of an inactivator of this invention.

Short letterhead of 1-p.qningpr

The invention shall be described in greater detail with reference to the attached drawings where: Fig. 1 is a diagram of percent residual activity of purified recombinant human ATase after incubation with different concentrations of different inactivators. Each point shows the average of 3 measurements. The line at 50% residual activity is used for calculating I^g values, i.e. the concentration of inactivator necessary to produce a 50% reduction in ATase activity. Fig. 2 are two graphs of percent cell growth versus alkylating agent concentration (ug/ml), showing the sensitizing effect of Q ar -benzylguanine (BeG) and Q C -tenylguanine (B.4205) at two different concentrations (0.1 and 1 ,0UM) on sensitization of Raji cells to BCNU. The line at 90% growth is used for calculating D^Q values, i.e. the dose of BCNU that produced 90% growth compared to untreated controls, i.e. 10% growth inhibition. Fig. 3 are four plots of percent cell growth versus alkylating agent concentration (PM), showing the effect of BeG and B.4205 at four different concentrations (0.1, 0.5, 1.0 and 5.0 UM) on sensitization of Raji- cells against temozolomide. Fig. 4 is a diagram extrapolated from fig. 3 of D,-g ( P- M) values (i.e., the dose of temozolimide that produced 50% growth compared to untreated controls) against the inactivator concentration (UM). Fig. 5 are two diagrams of percent cell growth against the inactivator concentration, showing the growth inhibition effect of Q<6->furfurylguanine (B.4203) and B.4205 on Chinese hamster V79 cells (RJKO) and a subline thereof (+120). Fig. 6 is two graphs of percent cell growth against the inactivator concentration, showing the growth inhibition effect of B.4203 and B.4205 on two Xernrtcrma pi gmpnro. siim subclones. Fig. 7 are four diagrams of percent cell growth against the inactivator concentration (UM) and which show the effect of degradation products of the inactivators B.4203, B.4205, B.4212 (Q-piperonylduanine) and B.4226

(Q -[2-benzo(b)thienylmethyl]guanine) on Raji cell growth.

Fig. 8 is a plot of ATase activity (fm/mg) versus time (hours) showing the depletion of ATase activity in A375 xenografts in nude mice for untreated controls, corn oil-treated controls, and extracts treated with BeG

(30 mg/kg) and B.4205 (30 mg/kg).

Figures 9-13 are diagrams of results of xenograft studies. In each figure, the upper graph shows percent tumor volume versus time (days) for A375 tumor xenografts in nude mice treated with BCNU alone, with BeG in combination with BCNU, and B.4205 in combination with BCNU. The lower graph in each figure shows the number of surviving mice in each treatment group versus time (days) after the treatments illustrated in the upper graph.

The details for each figure are as follows:

Fig. 14 is 3 graphs of percent survival versus temozolomide concentration (<U>M), showing the survival of bone marrow cells after treatment with inactivator (10UM) or DMSO (control) in combination with increasing doses of temozolomide.

Year of progress per for no f orp 1 sp. of oppf in nnpl sp. n

Q f-hetarylalkylguanine derivatives having the ATase-inactivating property can be synthesized by adapting the standard preparation presented below as appropriate.

2-amino-N,M,M-trimethyl-1H-purine-6-aminium chloride is prepared according to the method described by Kiburis et al., J.. Chem. Soc-( C .) , 1971, 3942. Details of the conditions for reaction of this quaternary salt with sodium benzyloxide (to give Q 6 -benzylguanine) not described in MacCoss, Chen and Tolman, T<p>1-rah<p> Mr. T.Ptr. r 1985, 26, 1815, was given in MacCoss, Tolman, Wagner and Hannah, European Patent Application No. 184,473, but these were not suitable for the preparation of relatively sensitive analogues, and the standard preparation below was designed by the inventors.

Standard preparation of Q_ g-hetarylalkylguanines

(formula I, y = H)

Sodium hydride (60% in oil; 0.8 g, 20 mmol) is added to a solution of RR<1>CHOH (56 mmol, approx. 5 ml) in DMSO (5 ml) and the mixture is stirred at room temperature for 1 hour. For solid or higher-molecular alcohols, up to 10 ml of DMSO can be used instead of 5 ml. 2-Amino-M,M,M-trimethyl-1H-purine-6-aminium chloride (2.29 g, 10 mmol) is added and stirring is continued for an additional 1 hour. The change in the UV spectrum is then complete (^^g 312->284 nm) and the almost clear solution is treated with acetic acid (1.7 ml). After cooling and diluting with ether (300 ml), the mixture is set aside for 2 hours and the solid (A) collected. Rubbing with water gives the product. A second fraction can be prepared by evaporating the ether-DMSO filtrate and trituring the residue successively with ether and water. Alternatively, the product can be extracted from the solid (A) with hot acetonitrile. The recrystallized compounds show a single spot in TLC (CgHg-MeOH,

4:1) and is characterized by analysis and its NMR spectra. Often they will contain solvent from the crystallization. Melting points and analytical data are given in Table 2, UV and H NMR data in Table 3. NMR spectra were measured on a Bruker WP80 or MSL 300 instrument.

This standard preparation procedure (with variations indicated by symbols in Table 2) was used to prepare the compounds listed in Tables 2a and 3a. In compounds B.4217 and B.4219, R' shall be methyl; in the remaining compounds R' shall be H. Q g -benzylguanine and the compounds B.4214, B.4218 and B.4231 listed in Table 4 below were also made by this standard preparation procedure for comparative testing purposes.

The variations indicated by the symbols in table 2 are as follows: a 5 mmol sodium hydride per mmol quaternary salt is used in the manufacture of this compound. When the standard amount (2 mmol) is used, 40% of the quaternary salt can be used

is recovered in the worked-up reaction.

h This compound is prepared by hydrolysis of the methyl ester B.4229 (145 mg, 0.5 mmol) in 2-methoxyethanol (2.5 ml) and water (2.5 ml) by treatment with 2M NaOH (2.5 ml) in 4 hours at room temperature. Neutralization with acetic acid (0.32 ml, 5.5 mmol), careful evaporation, trituration with water (3 ml) and filtration gives a solid which, on extraction with hot methanol, gives the acid B.4234.

a The required numbers are based on the monohydrate of a mixture of 4 parts of the sodium salt of the acid and 3 parts of

the acid, which requires Na, 4.3%. Found: Na, 4.44%.

d 3 mmol alcohol RCr^OH per mmol quaternary salt is used

instead of the standard 5.6 mmol,

b. For alcohols too sensitive to sodium hydride in DMSO at room temperature, prepare RCH20Na in DMF (2.5 ml

at -10°C; 3 mmol RCH 2 OH; 2 mmol sodium hydride). 1 mmol quaternary salt is added after 15-20 min. and stirring is continued for 2-3 hours at room temperature.

£ The products are extracted with acetonitrile.

Preparation of ribosides (formula I, Y = ribosyl)

A solution of the alkoxide made as in the standard procedure above from sodium hydride (60% in oil; 120 mg, 3 mmol) and RR'CHOH (4.6 mmol) in dry DMSO (2 mL) over 1 h is treated with 2- amino-6-chloropurine riboside (302 mg, 1 mmol) and stirred for 5 min. at room temperature, then for 15 min. at 60-65°C. The conversion is then complete as indicated by the change in the UV spectrum (max 3H-284 nm). Cooling and extensive trituration with ether (100 + 15 ml) and filtration gives a solid which is treated with water (10 ml). The pH is brought from 11 to 8 by briefly passing through CO2. Filtration removes inorganic material, and the dried residue is extracted at room temperature with methanol (4 x 10 ml portions, each containing one drop of pyridine). Evaporation of almost all the methanol and addition of some ether gives the product, almost pure by TLC (CgHg-MeOH 4:1).

It is recrystallized from methanol and a trace of pyridine, by dissolution and concentration below 40°C, sometimes with the final addition of a little ether.

This method was used to prepare the compounds listed in Tables 2b and 3b. Traces of contamination are difficult to remove, but NMR spectra show that the nucleosides are approx. 90% pure. The yields are of the order of 30-40%.

The starting alcohols RR'CHOH were usually made by reduction of the corresponding aldehydes, often commercially available, with sodium borohydride. A different method was used for the precursors of the ester B.4229 and the nitrile B.4273. Sucrose was converted''" vi a 5-chloromethylfurfural to 5-hydroxymethylfurfural. Oxidation of this aldehyde gave the carboxylic acid which was esterified by the method of Bocchi et al.<3> to methyl 5-(hydroxymethyl)furoate<4 >, required for B.422 9. The oxime c. of the aldehyde was dehydrated by the method of Carotti et al. g; the crude reaction product was treated with concentrated aqueous ammonia in methanol before extraction into dichloromethane. Distillation gave 5-cyanofurfuryl alcohol, required for B.4273.

For B.4266 Vilsmeier reaction Q of benzofuran gave the 2-aldehyde, reduced to the required alcohol Q, while for B.4226 lithiation and treatment with dimethylformamide gave the 2-aldehyde and in turn the alcohol 10

For B.4274, dimethyl tartrate was oxidized"<1>"<1> to methylglyoxylate which reacted 19 with tosylmethylisocyanide to form methyloxazole-5-carboxylate. This was reduced with lithium aluminum hydride by the method of Fallab14 to

the alcohol<15>.

For B.4275, bromomalonaldehyde gave 1 6 and thiourea, 2-aminothiazole-5-carboxaldehyde in 7 . Deamination with amyl nitrite 17 followed by sodium borohydride reduction gave 5-hydroxymethyl-

thiazole<14>.

For B.4271, 5-azapiperonyl alcohol (m.p. 82-84°C; found: C, 54, 60; H, 4.58; N, 9.09; C-^NC^ required: C, 54, 90; H, 4.61; N,

9.15%) prepared from the corresponding aldehyde18

For B.4278, 1-methylpyrrole-2-carboxaldehyde was nitrated 1 9 and reduced to the alcohol 2 0 with sodium borohydride.

As a specific example, the production of Q g-tenylguanine (B.4205) will now be described:

Preparation of Q g-tenylguanine

A solution of thenyl alcohol (3.18 mL, 33.6 mmol) in DMSO

(3 mL) was treated with sodium hydride (60% in oil; 0.48 g,

12 mmol), first by gentle stirring. After 1 hour, amino-N,N,N-trimethyl-1H-purine-6-aminium chloride (1.37 g, 6 mmol) was added and stirring was continued for another 1 hour. Acetic acid (1.0 mL) was added, cooled briefly, and the mixture diluted with ether (] mL). The solid (2.09 g) was collected after 1-2 hours. Evaporation of ether from the filtrate and distillation of DMSO and excess of thenyl alcohol (b.p. 48-57°C/0.2 mm) left a residue so trituration with ether gave a second solid fraction (0.36 g). The

combined solids were triturated with water (6 mL) to give the product (1.335 g, 90%) which showed a strong spot in TLC (CgHg-MeOH, 4:1) with only traces of contamination. Dissolve in hot ethanol (30 mL), clarify by filtration through celite, and concentrate (to 10 mL) using a rotary evaporator B.4205 (1.125 g, 71% material containing 1/3 EtOH per mole as solvate).

Compounds of formula I where Y is R"XCHR"<1> (seco-nucleoside can be prepared by an analogous preparation with the reaction of 0<6->benzylguanine with α-chloroethers (MacCoss et al., Tetrahedron Lett.; European patent application no . 184,473, loe. eit.) or with alkyl bromides (e.g. Kjellberg, Liljenberg and Johansson, Tetrahedron Lett., 1986, 27, 877; Moschel,

McDougall, Dolan, Stine and Pegg, J. Med. Chem., 1992, 35, 4486).

Typical "sugar" components corresponding to R"XCHR"<1>, leading to seco-nucleosides, are made by methods described in e.g. McCormick and McElhinney, J. Chem. Soc., Perkin Trans. 1, 1985, 93; Lucey, McCormick and McElhinney, J. Chem. Soc. Perkin Trans■ 1, 1990, 795.

Compounds of formula I where Y is ribosyl or deoxyribose (nucleosides) can be prepared by methods analogous to the syntheses of O<6->benzylguanine riboside and 2-deoxyriboside (Mochel et al., 1992; cf. Gao, Fathi, Gaffney. et al., J. Org. Chem., 1992, 57, 6954; Mochel, Hudgins and Dipple, J. Amer. Chem. Soc, 1981, 103 5489) (see preparation of ribosides above).

Industrial applicability

The amount of compound according to the present invention to be used will vary according to the effective amount required for the treatment of tumor cells. A suitable dosage is that which will result in a concentration of the compound according to the invention in the tumor cells to be treated, and which results in depletion of the ATase activity, e.g. about. 1 to 2000 mg/kg, and preferably 1 to 800 mg/kg, especially 1 to 120 mg/kg, before chemotherapy with an alkylating agent.

The pharmaceutical preparation according to the invention can be formulated in conventional forms with conventional excipients, as described e.g. in WO 91/13898, the contents of which are incorporated herein by reference. The preparation may contain the inactivator according to this invention together with an alkylating agent; or the preparation may comprise two parts, one containing the inactivator and the other containing the alkylating agent. The method for administering the compounds according to the invention to a host can also be a conventional method, as described e.g. in WO 91/13898.

For administration of an inactivator according to the invention to patients, the pharmaceutical preparation may conveniently contain the inactivator in a suitable carrier such as 40% polyethylene glycol 400 in saline solution, or in saline solution or 3% ethanol (in saline solution), for intravenous injection, or in powder form in suitable capsules for oral administration.

Alkylating agents can be administered according to known techniques and in conventional forms of administration, as described in e.g. WO 91/13898 or preferably as a single dose immediately after or up to 24 hours after, but preferably approx. 2 hours after administration of the ATase inactivating agents and also at doses lower than those used in standard treatment regimens. A reduction in dose may be necessary because the inactivators would generally be thought to increase the toxicity of the alkylating agents. Examples of chloroethylating agents include 1,3-bis-(2-chloroethyl)-1-nitrosourea (BCNU), 1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea (CCNU), fotemustine, mitozolomide and clomeson and those described at McCormick,

McElhinney, McMurry and Maxwell J. r. hpm. Snow. Pp. rlcin Trans. T, 1991, 877 and Bibby, Double, McCormick, McElhinney, Radacic, Pratesi and Dumont Anfilanrpr Drng nesign, 1993, 8, 115. Examples of methylating agents include temozolomide (British patent GB 2 104 522)

and dacarbazine, procarbazine, and streptozocin.

Methods

Purification of recombinant ATases

The cDNA cloning and overexpression of the human ATase has previously been reported 2 3. Purification of the recombinant protein* was achieved either by affinity chromatography through a DNA-cellulose column as described by Wilkinson et al. , 25 ' ? f, or 1 DEAE-cellulose ion-exchange chromatography. For the latter, the ATase protein was partially purified by ammonium

sulfate precipitation (30-60%) and dialyzed against 10 mM tris-HCl pH 7.5, mM DTT, 2 mM EDTA, 10% glycerol, before loading onto a DEAE cellulose column. The ATase was then eluted with a 0-0.1M NaCl gradient. The purified human ATase protein retained its activity for more than 1 year when stored at high concentration at -20°C in buffer I [50 mM Tris/HCl (pH 8.3)/3 mM dithiothreitol/1 mM EDTA ] and could be thawed and refrozen several times without significant loss of activity.

Incubation with inactivators and ATase assay

Compounds to be tested were dissolved in DMSO to a final concentration of 10 mM and diluted immediately before use in buffer I containing 1 mg/ml bovine serum albumin (IBSA). Recombinant ATase was diluted in IBSA and titrated so that reaction* could be carried out under ATase-limiting conditions, and not substrate-limiting conditions. In each assay, specific amounts of ATase (60-75 fmol) were incubated with varying amounts of O<6->

benzylguanine, or test compound in a total volume of 200 ul

IBSA containing 10 µg calf thymus DNA at 37°C for 1 hour. The [<3>methylated DNA substrate (100 µl containing 6.7 µg DNA and 100 fmo!

of 0 -methylguanine) was added, and the incubation continued at 37° for an hour, until the reaction was complete. After acid hydrolysis of I

as previously described 21 the [ 3 H]-methylated protein was recovered* and quantified by liquid scintillation counting. The samples were typically analyzed in duplicate and the experiments repeated several times. I^q is the concentration of inactivator necessary to produce a 50% reduction in ATase activity.

Cell culture and preparation of extracts

Mammalian cells were cultured under standard conditions. E.g. Raji (a human lymphoblastoid cell line from a Burkitt's lymphoma) cells were grown in suspension culture in RPMI medium supplemented with 10% horse serum. A375M cells are human melanoma cells from which the xenografts described below were established after subcutaneous injection into nude mice: WiDr cells are a human colon carcinoma cell line: Hamster+120 cells are a mitozolomide-selected subline of a Chinese hamster lung fibroblast V79- cell line called RJKO: Yoshida cells are a rat carcinosarcoma cell line and YRbus is a busulphan resistant subline thereof: XP cells are an SV40-transformed fibroblast cell line originally from the skin of an Xpmdprma pi m& ntnsnm patient, pHMGhAT2b cells are a clone of those cells that have been transfected with a mammalian cell expression vector containing the human ATase cDNA and pHMGla cells are a clone that has been transfected exclusively with the expression vector (i.e., one that does not contain the human ATase cDNA). Cell pellets were resuspended in cold (4°C) buffer I containing 2 µg/ml leupeptin and sonicated for 10 sec. at a distance between the peaks of 12 Pm. After cooling in ice, the cells were sonicated for a further 10 sec. at 6 p.m. Immediately after the ultrasound treatment, 0.01 volume of 8.7 mg/ml phenylmethanesulfonyl fluoride in ethanol was added, and the ultrasound-treated cells centrifuged at 15,000 G for 10 min. at 4°C to pellet cell waste. The supernatant was saved for determination of ATase activity (see below).

Stability of inactivators at 37°C.

Inactivators (10 mM in DMSO) were diluted to 0.1 mM in prewarmed degassed buffer I (1 mM EDTA, 50 mM tris pH 8.3) or PBS (pH 7-7.2). PBS (phosphate buffered saline) is 0.8% NaCl, 0.02% KCl, 0.15% Na2 HP04, 0.02% KH2PO4, pH 7.2. The samples were immediately transferred to the CARYl3 spectrophotometer (cuvette block kept at 37°C) and scanned at an appropriate wavelength in accordance with

the compound's spectral properties with 3-10 min. space in up t;

80 hours. The results were expressed as percent absorbance change versus time and Tl/2 values (half-life) extrapolated from the

Inactivation of ATase activity in mammalian cells

Cells were diluted to 10 3 /ml in culture medium containing either the appropriate concentration of inactivator or an equivalent volume of vehicle (DMSO). After incubation at 37°C for 2 hours, the cells were harvested by centrifugation, washed twice with PBS and the resulting cell pellets (between 1-2 x 10 per pellet]

stored at -20°C. The ATase activity was determined as described

above, in duplicate ultrasound-treated cell extracts and expression? as percent residual activity based on the activity present in untreated controls (eg 350-450 fm/mg in Raji cells). values (i.e., the concentration of inactivator required to reduce ATase activity by 50%) were extrapolated from these data.

Sensitization of Raji and A375M cells to BCNU and

temozolomide

Sensitization of Raji cells to the cytotoxic effects of BCNU and temozolomide after 2 hours of pretreatment with inactivator l

22

analyzed using an XTT assay. Briefly, the cell* were plated at 1000 cells/well in 96-well plates and incubated

37°C for 30 min. before addition of medium containing either the appropriate concentration of inactivator or an equivalent volume of carrier. After 2 hours of incubation at 37°C, medium containing either increasing doses of BCNU, temozolomide or an equivalent carrier was added, and the cells were allowed to grow for 6 days. At this point, XTT solution was added and the cells incubated further

4 hours at 37°C. The resulting red/orange formazan reaction product was quantified by measuring the absorbance at 450 nm on a microtiter plate reader. Sensitization of A375M cells to the cytotoxic effects of BCNU was analyzed by the MTT assay<24> which differs from the XTT assay described above as follows. A375M cells (1000 per well) were allowed to grow for 24 hours and then treated with the inactivator, and 2 hours later with BCNU. After 6 days, MTT solution (4 mg/ml) was added and the cells incubated for 3 hours at 37°C. Medium was aspirated and the resulting purple colored formazan crystals were solubilized in DMSO (120 µl) and cell viability was quantified by measuring the absorbance at 540 nm using a microtiter plate reader. From these data, the percentage growth of cells relative to the growth in control wells was determined for a range of BCNU or temozolomide doses in both the presence and absence of inactivator. Raji sensitization to BCNU (DgQ.<C/>D9Q.<I>) was determined by dividing D 90 (i.e., the dose at which there was 90% growth versus untreated controls, i.e., 10% growth inhibition) calculated for data using BCNU alone (Dgo. p) with the growth for BCNU plus inactivator (DgQ.<1>). A value of one (1) thus indicates no sensitization with the inactivator. Raji sensitization to temozolomide and A375M sensitization to BCNU were determined using the corresponding D5Q values (ie, the doses at which there was 50% growth inhibition).

Sensitivity of mammalian cells to the inactivators or their hydrolysis products

To determine the effect on cell growth alone was

Xeroderma pigmentosum cells and Chinese hamster V79 cells exposed to increasing concentrations (up to 600uM) of selected inactivators for 2 hours at 37°. In some cases, the inactivators were allowed to undergo hydrolysis at 37°C for 20 hours and then added to Raji cells to determine the extent to which the decomposition products of the agents could inhibit cell growth. After 6 days, the degree of cell growth was determined as described above.

Sensitization of bone marrow cells to temozolomide (GM-CFC assay)

For the granulocyte/macrophage colony-forming cell (GM-CFC) assay, primary human bone marrow samples were obtained from patients undergoing cardiothoracic surgery. After removal of erythrocytes J

the samples, the cells were plated at 1-2 x 10<5>/ml in 300mOsM Iscove's medium containing 10uM inactivator or equivalent volume of DMSO, 20% fetal calf serum, 10% 5637 conditioned medium as a wedge for growth factors and 0.3% agar nobel , in 1 ml Petri dishes containing appropriate doses of temozolomide and incubated at 37°C in an atmosphere of 5% CO 2 and 95% air. After 9 days, colonies comprising more than 50 cells were counted. The colonies represented progenitor cells for the granulocyte/macrophage lines huGf CFC. Survival was expressed as a percentage of the number of colonies at zero dose of temozolomide.

Xenograft studies

Animals

BALB-C-derived male athymic mice (nu/nu athymic) weighing between 25 and 35 g were received from the breeding colony at the Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Wilms! Road, Manchester M20 9BX, England (ASU mouse). The animals were kept in a sterile environment until they were needed for experiments. In one experiment, the mice were supplied from Harlan-Olac, Harlan UK Limited, Shaw's Farm, Blackthorn, Bicester, Oxon. 0X6 OPT, England (OLAC mouse). These mice weighed between 17 and 23 g and were subsequently found to be more sensitive to the toxic effects of the combination (inactivator and BCNU) treatment. For this reason, lower doses were administered.

Cells

A375M cells (human melanoma) were grown in DMEM containing 10% fetal bovine serum. The cells were prepared for ±n vi vn inoculation by incubation with 0.01% trypsin.

Tumors

Cells (10<6>) in 100 µl PBS were injected subcutaneously into the right hand flank of 8-10 week old nu/nu athymic mice. These cells were allowed to develop into a tumor for 3-4 weeks before transfer to experimental animals. Tumor blocks measuring 2 mm x 2 mm x 2 mm were implanted subcutaneously on the flank of the right hand. These animals were ready for use in inactivator experiments in 7-10 days.

Drug treatment

Nu/nu mice were treated with either 30 or 60 mg/kg Q6-benzylguanine, or B.4205 and the appropriate vehicle control (i.p), 60-90 min. before 10-25 mg/kg BCNU (i.p). Q<6->benzylguanine and B.4205 were prepared as a 5 mg/kg solution in corn oil and BCNU (2 mg/ml) in PBS + 3% ethanol.

Tumor measurements

Xenograft tumor measurements were taken every 1–2 days by workers using digital chrome calipers. Tumor volume was calculated using the formula (h x w x 1) <n>/6. The experimental animals were also weighed every 1-2 days. The measurements continued until the tumors in the control animals reached the maximum allowed volume (i.e. 1 cm x 1 cm x 1 cm).

Results

Tables 4, 5 and 6 show the physical, biochemical and inv<_>Lv_Q data for each of the inactivators. The listed tests are explained in the Methods section.

Fig. 1 shows the result of in vitro trn ATase inactivation assay using 4 of the compounds. B.4206 was somewhat more effective than BeG, but B.4212 and B.4205 were significantly better under the assay conditions used. The B.4203 was not as effective as the BeG. Fig. 2 shows that 0.1 UM inactivator, B.4205 was more effective than BeG in sensitizing Raji cells to the growth-inhibitory effects of BCNU: at 1.0 UM inactivator, B.4205 and BeG were equally effective in this respect. Fig. 3 shows that 0.1UM inactivator, B.4205 was more effective than BeG in sensitizing Raji cells to the growth-inhibitory effects of temozolomide, but that as the doses of the inactivators were increased, sensitization by BeG became more effective, while the one at B.4205 remained the same. This lack of dose response with B.4205 but clear dose response with BeG is shown more clearly in fig. 4. Fig. 5 shows that while some growth inhibition of B79 cells was produced by B.4203 at doses in excess of 100 UM (i.e. at least 100x higher than the I^g dose for this compound see), no such effects were seen with B.4205 up to the maximum concentration used. Fig. 6 shows that XP cells were equally susceptible to the growth-inhibitory effects of B.4203, but that these cells were more sensitive than V79 cells to the effects of B.4205. However, the doses required for growth inhibition were at least 100x those required for ATase inactivation. It can be concluded that the inherent growth-inhibitory effects of these inactivators would not demonstrably contribute to the sensitization of cells to the growth-inhibitory effects of the alkylating agents. Fig. 7 shows that no significant growth-inhibitory

effects were produced in Raji cells when the given compounds were allowed to undergo hydrolysis (see methods) before being added to the cells without further addition of alkylating agents. Under these experimental conditions, the decomposition products of the agents would therefore not be expected to contribute to the growth inhibitions observed when using combinations of

these agents and alkylating agents.

Fig. 8-13 shows the results of the xenograft study in greater detail. The depletion and recovery of ATase activity after exposure to Q. -benzylguanine or B.4205 was measured in A375 tumor xenograft extracts (Fig. 8) prepared from tissues of animals sacrificed 2 or 24 hours after inactivator administration. Fig. 9-13 indicates the sensitization of A375 tumor xenografts in nude mice to either Q -benzylguanine or B.4205 in combination with BCNU as described in the Methods section. In each figure, the upper diagram should show the percentage increase in tumor volume over the course of the experiment. The lower diagram shows the number of animals in each treatment group, and the number of survivors after treatment. The diagrams show that B.4205 is comparable to Q -benzylguanine (BeG) in reducing tumor volume, and is significantly less toxic in combination with BCNU than BeG in combination with BCNU. In human patients, cutaneous malignant melanoma is treated with BCNU, particularly in the USA (CM. Balch, A. Houghton & L. Peters (1989) "Cutaneous Melanoma", in Cancer: Prinr. iplps and Pra.tis<p> of Onr-.nl ogy, V.T. De Vita, S. Helman & S.A. Rosenberg, (Eds.), pp. 1499-1542, Lipincott: Philadelphia) so that the human melanoma xenograft cultured in nude mice is an animal model system that is clinically most relevant.

Fig. 14 shows the results of the tests of sensitization of human bone marrow cells to temozolomide as described in the Methods section above. The bone marrow cells were supplied from three patients identified as C, D and E respectively. The survival curves shown relate to the cytotoxic effect of combined treatment with inactivator/temozolomide. It is desirable that this effect should be reduced. The results for patients C and E show that B.4205 had a smaller sensitizing effect than Q -benzylguanine (Q BeG); for patient D there was no difference, but this result is considered a reasonable variation in scientific testing with human material.

Table 7 shows the toxic effect of inactivator alone on bone marrow cells delivered from 5 patients A-E, including the same patients C-E as in fig. 17. The results for B.4205 are comparable to those for 0. -benzylguanine. Note that the cells from patient B were almost twice as sensitive to both BeG and B.4205 compared to the other samples.

Summary of the findings

1) All compounds have been tested for their ability to inactivate recombinant human ATase under standard conditions in an in vJ_txo assay. Under these conditions, two of the compounds B.4214 and B.4217 would not inactivate ATase up to the highest concentration used, but these were compounds where R<*> is methyl. The rest of the compounds inactivated ATase with I^Q values ranging from 0.0045 to 95 Umolar. 8 compounds (B.4205, B.4206, B.4212, B.4226, B.4266, B.4269, B.4273, B.4275) had I5Q values lower than that of BeG

(Table 4).

2) The inactivators underwent hydrolysis in aqueous solution at different rates (half-lives 0.17 to >80 h) but this was not related to their effectiveness as ATase inactivators (Table 4). 3) Compounds that were effective in the inactivation of ATase in vitro (I50 <1/0UM) also inactivated ATase in human cells with I^g values that were generally only slightly higher (on average about 1.5x) than those that were found using recombinant protein in ±n iitxa assay (Table 4). 4) B.4205 inactivated ATase in human cells, rat cells and Chinese hamster cells with similar efficiency (I^g values 0.02-0.03). For B.4203, the range was 0.04-0.12. In the cell lines used, B.4205 and B.4203 were up to 7x more effective than BeG (table 5). 5) B.4203 and B.4205 were toxic to the studied XP cells, and B.4203 was toxic to the studied V79 cells, but only at concentrations that were at least 100x higher than those at which sensitization to BCNU was observed (fig .5 and 6). 6) Compounds that were effective in the inactivation of ATase in vitro (i5Q <1.0UM) and in Raji cells (I5Q <<1>.0PM) would also sensitize Raji and A375M cells to the growth-inhibitory effects of BCNU where this was tested (table 4) . 7) At inactivator concentrations of 0.1 UM, B.4205 was more effective than BeG in sensitizing Raji cells to the growth-inhibitory" effects of temozolomide (Fig. 3). 8) the in vj_tra assay can be used to predict which compounds which are most likely to be effective sensitizers of mammalian cells to the growth-inhibitory effects of BCNU and related cytotoxic agents.9) B.4205 was similar or slightly more effective than BeG in sensitizing human melanoma xenografts grown in nude mice to the growth- inhibitory effects of BCNU (Figs. 9-13). 10) B.4205 was as effective as BeG in inactivating ATase in human melanoma xenografts grown in nude mice (Fig. 8). 11) ln the vj_txni assay and/or The xenograft ATase depletion assay can be used to predict which compounds are most likely to be effective sensitizers of melanoma xenografts to the growth-inhibitory effects of BCNU and related cytotoxic agents.12) In contrast to BeG, which caused caused death in up to 70% of the treated animals, B.4205 had very little effect on the sensitivity of nude mice with human melanoma xenografts to the acute toxic effects of BCNU under the conditions used

(fig. 9-13).

13) BeG sensitized GM-CFC in the three human bone marrow samples tested for the toxic effects of temozolomide, but in two of these samples little or no sensitization was produced by B.4205. This assay can therefore be used to predict the possible myelosuppressive effects of ATase inactivators when they are used in the clinic in combination with BCNU and related agents (Fig. 14).

R<p>f<p>.r^nfi<p>r

1. W.N. Haworth and W.G.M. Jones, J. rhpm. No. pp. 1944, 667.

2. T. Reichstein, Heiv.Chim.Acta, 9, 1926, 1066.

3. V. Bocchi, G. Casnati, A. Dossena and R. Marchelli,

SynthPsis, 1979, 961.

4. P.A. Finance, J- Ch<p>m. No. born 1963, 3917; YES. Moore and J.E.

Kelly, J.Polym.Sei., Polym.Chem.Ed., 22, 1984, 863.

5. J. Kiermayer, r.hpmi Vpr-7, Pi tung, 19, 1895, 1004.

6. A. Carotti, F. Campagna and R. Ballani, Synhe. say p, 1979,

56. 7. A.J. Floyd, R.G. Kinsman, Y. Roshan-Ali, and D.W. Brown,

TPtr^h<p>rirnn, 39, 1983, 3881.

8. V.T. Suu, N.P. Buu-Hoi and N.D. Xuong,

Rnl 1. No. rhin. FranrP, 1962, 1875.

9. T. Reichstein and L. Reichstein Hel v. Chim. Acta, 13, 1930, 1275.

10. H. Takeshita, H. Mametsuka and H. Motomura,

J. HPtPrnr. yrl Chim. f 23, 1986, 1211.

11. F.J. Wolf and J. Weijlard, no. Synth. r r . nl in Vnl .& r 1963,

124 . 12. A.M. van Leusen, B.E. Hoogenboom and H. Siderius,

TPtrah<p>Hrnn Lett., 1972, 2369.

13. A. Maquestiau, R. Flammang and F.B. Abdelouahab,

HptPrnryrlpsf 29, 1989, 103.

14. S. Fallab, Hplv. nhim. Arta, 35, 1952, 215.

15. A.S. Kende, K. Kawamura and R.J. DeVita, J. Ampr. chem. Sn r., 112, 1990, 4070.

16. S. Trofimenko, J. no. rhpm f 28, 1963, 3243.

17. I. Sawhney and J.R.H. Wilson (Shell International),

EP.395.174, 1989, ( rhpm. Ahs , 114, 143410s).

18. F. Dallacker, P. Fechter and V. Mues, Z. Natnrfnrsrh.,

34b, 1979, 1729.

19. P. Fournari, Bnl 1 . finr ; . Chim. Fran no. P, 1963,488.

20. L. Grehn, P. hPm. Srr. b 16, 1980, 72.

21. Morten, J.E.N. & Margison, G.P. (1988) Carcinogenesis 9, 45-49. 22. Scudiero, D.A. Shoemaker R.J., Pauli K.D., Monks A.,

Tierney S., Nofziger T.H., Currens M.J., Seniff D. & Boyd

M. R. (1988), Cancer Research 48, 4827-4833.

23) Fan, C-Y, Potter, P.M., Rafferty, J.A., Watson, A.J., Cawkwell, L., Searle, P.F., 0'Connor, P.J. and Margison,

GP (1991) Nucleic Acids Res. 18, 5723-5727.

24) Carmichael J., DeGraff W.G., Gazdar A.F., Minna J.D., Mitchell J.B., (1987). Evaluation of a tetrazolium based semiautomated colorimetric assay : assessment of chemo-sensitivity testing. Cancer Res. Al: 936-942. 25) Wilkinson, M.C, Potter, P.M., Cawkwell, L., Georgiadis, P., Patel, D., Swann, P.F. and Margison, G.P. (1989)

Nucleic Acids Res. 17, 8475-8484.

26) Wilkinson, M.C, Cooper, D.P., Southan, C, Potter, P.M. & Margison, G.P. (1990) Nucleic Acids Res., 18, 13-16.

In the description, the abbreviations "lh" or "2h" etc. mean "1 hour", "2-hour" etc.

Claims (18)

1. Q-alkylguanine derivatives, characterized by the formula I: where Y is H, ribosyl, R<1> is H or C1-C5 alkyl, R is (i) . a cyclic group with one 5- or 6-membered heterocyclic ring, optionally with a 6-membered carbocyclic or 5- or 6-membered heterocyclic ring fused thereto, or wherein the one or each heterocyclic ring has at least one heteroatom selected from 0, N or S, or a substituted derivative thereof with substitution in the heterocyclic ring(s) and/or carbocyclic ring(s) with one or more of halogen, cyano or S0nR where R"" is C1-C5-alkyl and n = 0, 1 or 2; or (ii) naphthyl and pharmaceutically acceptable salts thereof. 2. Connection according to claim 1, characterized in that R is: (a) a 5- or 6-membered heterocyclic ring, or (b) a benzo derivative thereof, the Q -alkylguanine group being bound to R in either the heterocyclic or benzene ring. 3. Compound according to claim 1 or 2, characterized in that R is a 5-membered hetero cyclic ring that has at least one S atom in it. 4. Compound according to claims 1-3, characterized in that R is selected from a thiophene ring, a furan ring or a substituted derivative thereof as defined in claim 1. 5. Connection according to claim 1, characterized in that R includes a heterocyclic ring substituted with halogen or cyano. 6. Compound according to any one of claims 1-5, characterized in that R is selected from a thiophene ring, a furan ring and substituted derivatives thereof selected from bromine- and cyano-substituted derivatives thereof. 7. Compound according to claim 6, characterized in that -CHRR' is tenyl or a bromine-substituted derivative thereof. 8. Connection according to claim 7, characterized in that it is Q<6->thenylguanine or a bromine-substituted derivative thereof. 9. Connection according to claim 1, characterized in that R' is H. 10. Connection according to claim 1, characterized in that it is selected from fi f f Q. -thenylguanine, Q - (3-thienylmethyl) guanine, Q. - piperonylguanine, Q<6->furfuryl-guanine, Q<6->(2-benzo[b]thienylmethyl)guanine, Q c-(2-benzofuranylmethyl)guanine, Q - (5-thiazolylmethyl)guanine, Q. - ( bromothenyl)guanine, Q -(5-cyanofurfuryl)guanine and Q<6->(2-naphthylmethyl)guanine. 11. Pharmaceutical preparation, characterized in that it comprises a compound according to any of the preceding claims, and a pharmaceutically acceptable excipient. 12. Pharmaceutical preparation according to claim 11, characterized in that it further comprises an alkylating agent. 13. Compound according to any one of claims 1-10, characterized in that it is to be used in a method for depleting Q-alkylguanine-DNA alkyltransferase activity in a host. 14. Use of a compound according to any one of claims 1-10, for the production of a drug for depleting Q_ -alkylguanine-DNA alkyl-transferase activity in tumor cells. 15. Use of a compound according to any one of claims 1-10 and an alkylating agent for the production of a one- or two-part pharmaceutical preparation for the treatment of tumor cells. 16. Use according to claim 14 or 15, where the compound according to any one of claims 1-10 is Q 2 -thenylguanine or a bromine-substituted derivative thereof. 17. Preparation according to claim 12, characterized in that the alkylating agent is selected from 1,3-bis-(2-chloroethyl)-1-nitrosourea (BCNU) and temozolomide. 18. Process for producing a compound of formula I according to claim 1, characterized in that it comprises reaction of sodium hydride with a solution of RR'CHOH (where R and R' are as defined in claim 1) in an organic solvent, addition of 2-amino-N,M,N-trimethyl-1H-purine-6-aminium chloride or 2-amino-6-chloropurine riboside, treatment with a weak acid and ether, and extraction of the desired product.